Photovoltaic technology has been developed under a background of global warming and exhausting of fossil fuels. The substitutes of traditional energy sources include nuclear energy and renewable energy. Among them, the nuclear energy will also be exhausted in the future. Moreover, the potential radiation contamination arising from the nuclear energy may bring serious problems to our environment, especially after the recent accident in a nuclear power station of Japan. Therefore, our future may greatly rely on renewable energy. The photovoltaic devices, or solar cells, are playing a leading role in the renewable energy. This enormous future demand has dramatically pushed the development of photovoltaic technology. At present, the second generation photovoltaic devices, thin film solar cells, have appeared in the global market. They currently include three main types: amorphous silicon, CIGS and CdTe. In this thin film solar cell family, an amorphous silicon cell has a low conversion efficiency that may reach up to 13% for a triple junction design. Besides, it has a problem of power degradation with initial illumination, but its technology is relatively mature. By contrast, a CIGS solar cell possesses the highest conversion efficiency that is as high as 20%, higher than 17% efficiency of the CdTe ones. In the periodic table of the elements, the elements of a CIGS absorber are located in Group IB-IIIA-VIA and the ones of a CdTe absorber in Group IIB-VIA. These absorber materials all belong to multi-component p-type semiconductors. For such a semiconductor material, the distribution of different components and stoichiometry may determine the quality of the material.
Both of CIGS and CdTe solar cells contain a stack of absorber/buffer thin film layers to create an efficient photovoltaic heterojunction. A metal oxide window containing a highly resistive layer, which has a band gap to transmit the sunlight to the absorber/buffer interface, and a lowly resistive layer to minimize the resistive losses and provide electric contacts, is deposited onto the absorber/buffer surface. This kind of design significantly reduces the charge carrier recombination in the window layer and/or in the window/buffer interface because most of the charge carrier generation and separation are localized within the absorber layer. In general, CIGS solar cell is a typical case in Group IB-IIIA-VIA compound semiconductors comprising some of the Group IB (Cu, Ag, Au), Group IIIA (B, Al, Ga, In, Tl) and Group VIA (O, S, Se, Te, Po) elements of the periodic table. These elements are excellent absorber materials for thin film solar cells. In particular, compounds containing Cu, In, Ga, Se and S are generally referred to as CIGS(S), or Cu(In,Ga)(S,Se)2 or CuIn1-xGax (SySe1-y)n, where 0≦x≦1, 0≦y≦1 and n is approximately 2, and have already been applied in the solar cell structures that gave rise to conversion efficiencies approaching 20%. Here, Cu(In,Ga)(S,Se)2 means the whole family of compounds with Ga/(Ga+In) molar ratio varying from 0 to 1, and Se/(Se+S) molar ratio varying from 0 to 1. It should be noted that the molar ratios of Ga/(Ga+In) and Cu/(Ga+In) are very important factors to determine the compositions and the conversion efficiencies of the CIGS solar cells. In general, a good CIGS solar cell requires a ratio of Cu/(Ga+In) between 0.75 and 0.95, and Ga/(Ga+In) between 0.3 and 0.6. In comparison with CIGS, the composition of a CdTe solar cell is much simple. In general, the content of Cd is close to 50% in the CdTe films. However, the Cd content may change after the deposition of a CdS layer and the subsequent annealing procedure. Close to the interface of the p-n-junction, for example, a CdSxTe1-x layer is formed with x usually not exceeding 0.06. However, x has a range changing from 0 to 1, which results in a compound from CdTe (x=0) to CdS (x=1).
Both CIGS and CdTe films have to be annealed to form a uniform stoichiometric compound. A CIGS film is usually annealed at a temperature between 350 and 600° C. in a typical two-stage fabrication procedure. For a CdTe solar cell, a CdS film may firstly be annealed in a superstrate configuration and a CdS/CdTe bilayer may be annealed in a substrate configuration. After annealing, an n-type semiconductor buffer layer such as CdS, ZnS, or In2S3 should be deposited onto a CIGS semiconductor absorber. After then, transparent conductive oxide (TCO) materials, i.e., ZnO, SnO2, and ITO (indium-tin-oxide), should be deposited to form the solar cells. For a CdTe solar cell, CdS may be deposited onto the surface of CdTe or TCO, depending on its substrate or superstrate configuration. However, a superstrate configuration may not be applicable on a flexible workpiece on the basis of current technology. Therefore, CdS should be deposited onto a CdTe surface to fabricate a CdTe thin film solar cell on a flexible substrate. The CdS thickness requirement is similar to a CIGS and a CdTe cell. They usually require a thin layer of CdS with about 100 nm or thinner.
There are some different technologies to deposit CdS, such as vacuum sputtering and evaporation, spray pyrolysis and chemical bath deposition. Among them, the chemical bath deposition (CBD) seems to produce the best result. In comparison with the vacuum methods, the CBD process is simple and requires cheap equipments. The disadvantage is that it produces lots of cadmium wastes and the post-treatment may be a heavy work.
A common CdS chemical deposition bath contains a cadmium salt and a thiourea solution in ammonia medium. The ammonia medium in the solution plays two roles. On one hand, it provides OH− ions to hydrolyze thiourea, which slowly releases S2− into the solution. On the other hand, it controls the amount of Cd2+ cations in the bath through the generation of its tetra-ammonium complexes that decompose to slowly release Cd2+ into the solution. The slowly created Cd2+ and S2− ions synthesize the CdS, some of which deposits onto the substrate surfaces. The reaction mechanism may be shown below:a. [Cd(NH3)4]2+=Cd2++4NH3 the instability constant of [Cd(NH3)4]2+:Ki=7.56×10−8.  (1)b. (NH2)2CS+2OH−→CH2N2+2H2O+S2−  (2)c. Cd2++S2−→CdS  (3)The stability product of CdS is: Ksp=1.4×10−29 in a strong alkaline solution.There is a possible side reaction below:Cd2++2OH−→Cd(OH)2  (4)The solubility product constant of Cd(OH)2 is: Ksp=7.2×10−15. However, Cd(OH)2 will dissolve in the ammonia alkaline solution to become [Cd(NH3)4]2+ complex cations and not affect the quality of the CdS product.
The CdS precipitation can take place either in the bulk solution to form colloids or at the immersed substrate surface to generate a layer of thin film. According to a main theory, the layer deposition may take three stages: 1). Induction with the ion adsorption and formation of nucleation centers; 2) layer growth with an “ion by ion” mechanism; and 3) layer growth with a “cluster by cluster” mechanism.
The key technology is that the solution must be freshly prepared to make the initial nucleation of CdS to take place directly on the surface of a substrate. Otherwise, the CdS deposition on the substrate surface may be powdery because the initial induction mainly takes place in the bulk solution and the deposition on the substrate may be directly from the “ion-by-ion” or even “cluster-by-cluster” stages. As reported in a U.S. Pat. No. 7,846,489 B2, which is incorporated herein by reference, the formation of CdS in a reaction solution follows two directions, i.e., a homogeneous particle formation in the solution and a heterogeneous film growth on a substrate surface. No matter which kind of mechanism is involved in the CdS formation in an alkaline solution, i.e., ion-by-ion, hydroxide cluster, or complex-decomposition mechanism, as discussed in literatures, a uniform growth of a CdS thin film on a substrate requires interaction between the substrate surface and un-reacted species firstly. A freshly prepared solution can meet this requirement very well. The reaction starting from a freshly prepared solution is not difficult to achieve in a laboratory work. The sample can be simply immersed into a freshly mixed solution at a certain reaction temperature. In fact, there are lots of literature reports relating to the CdS chemical bath deposition in laboratory works, including some patents such as U.S. Pat. No. 6,537,845 B1 describing a CdS chemical surface deposition onto glass substrates coated with CIGS films. However, it is difficult in an industrial manufacture process, especially for a continuously traveling flexible substrate in a roll-to-roll or reel-to-reel process. How to freshly mix a reaction solution on a moving flexible substrate and remain this mixed solution on the substrate surface to react for a period of time at a constant enhanced temperature is a great challenge for a manufacture design.
Although there are plenty of patents and publications related to CdS deposition via CBD methods, the apparatus or equipments for industrially manufacturing CdS or ZnS buffer layers in solar cells with a CBD process are seldom reported. Recently, some CBD apparatus and methods were described in U.S. Pat. No. 7,541,067 B2, U.S. Pat. Appl. Pub. Nos. 2011/0039366 A1, 2009/0255461 A1 and 2009/0246908 A1, which are incorporated herein by references. In these patent and applications, the CdS deposition apparatus were designed for roll-to-roll manufacture processes in fabrications of thin film CIGS solar cells. In these apparatuses, the freshly mixed solutions were delivered onto the surfaces of travelling flexible substrates and remained there during the periods of deposition reactions to grow up the CdS thin films.
The apparatuses presented in the patents described above require some special designs to remain the solutions on the surfaces of the moving flexible substrates in a roll-to-roll process. For example, the substrates might need to be bent using some magnetic or mechanical forces to hold the solutions on the tops. These applied forces may affect smooth movement of the flexible substrates. It might be difficult to remain the surfaces flat, especially for a wide web. It might always be a great challenge to remain backside of the substrates drying. The present inventor has submitted a patent application with an application Ser. No. 13/154481 to solve these problems. The apparatus presented in that application does not have to bend the flexible substrates and fit to different widths. As a divisional application, the present invention provides a spray method to carry out a CBD reaction with the apparatus invented in the parent application Ser. No. 13/154481.